
slinging tokens on @aiatmeta to help businesses grow
USA
Joined July 2009
- Tweets5.4K
- Following866
- Followers19.1K
- Likes11.8K
Jack Krawczyk retweeted
Selling something on Marketplace or running a small business? Muse can now turn your photos into polished, professional-looking images for your listing or your website. No studio needed.
Jack Krawczyk retweeted
Today we’re publishing the first version of Poppy, an open standard we’re building with industry partners for how personal agents work with businesses. We’re proud that Muse is the first personal agent to support it. More on Poppy and how businesses can get started: muse.ai/blog/poppy-personal-…
Nathan is an incredible partner for helping the team and me better learn how to serve businesses with @muse.
He's a renaissance man and Army veteran who has been pushing the envelope on the best ways to grow and create jobs for his community.
His story: facebook.com/business/news/f…
Our military deserves an ovation every day for their service. For over 250 years, they have protected the freedom that makes America the land of opportunity, the land of the free.
I am grateful to every man and woman who have served our nation. We would not have an ounce of the prosperity we experience without you.
God bless our veterans and God bless America 🇺🇸
5
624
my favorite use case: on a recent trip to italy & france, my muse emailed and coordinated reservations with local restaurants who don’t list tables online
10/10. would recommend to a friend.
Jack Krawczyk retweeted
use muse ember to save millions
pay $0.10 / $0.50 / Mtok and save millions of dollars on legal clerical errors
We put Muse Ember from @AIatMeta to work untangling conflicting records in a $12.6 million private-fund capital call.
Across 13 documents in Box, Ember catches a $2.48 million misallocation and a missed $250,000 credit, recalculates each investor’s obligation, and produces a source-cited review for Fund Operations.
Customers can build custom agents for their own document-heavy financial workflows with Box AI Studio.
Jack Krawczyk retweeted
Releasing Muse Ember today -- our smallest model that delivers higher intelligence per dollar than others in its class. Built for token-heavy work that agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing. Available today on the Meta Model API and in Muse Code.
Jack Krawczyk retweeted
Upgraded the models for Muse and Meta Al, bringing Active Memory and Active Focus from Muse Wonder to everyone. In Muse, these new specialized models are 2.4x less likely to forget critical info and are 67% better at self correcting and completing everyday tasks which means less repeating yourself and fewer dead ends.
More on the models: research.meta.ai/blog/specia…
it's been great feedback so far and more room to keep getting better muse.ai/business
Muse for SMBs might be the most underrated story in consumer agents right now
They have near-infinite consumer distribution at very low cost, with a base of 3.64 billion daily active users across Meta apps
But even more importantly, more than 500M SMBs use Meta for marketing. If Meta gives agent discoverability (preferred or not) to that SMB base, it could be a huge boost, especially for local businesses who can hyper-target users with urgent purchase needs
Meta said in their Q3 earnings calls (and in blog posts prior) that they plan to build an ecosystem of connectors for SMBs in Muse - analytics across your Meta accounts plus QuickBooks, Canva, Stripe, etc.
Many of these businesses are already trusting Meta with their growth / ad dollars - I could see Muse becoming almost a nervous system for SMBs that allows Meta to capture more and more of their activity over time
Shopify has already leaned into this (@tobi) - Muse can search across the Shopify catalog and complete purchases for consumers, and businesses can manage their Shopify store in Muse via a connector
And, the early Muse user base has shown a willingness to purchase via agents. 25% have completed a transaction via Muse, with average purchase size over $200. And, the top activity on Muse is shopping
Very curious to see what comes next here, especially with holiday gifting on the horizon 👀
Jack Krawczyk retweeted
konnichiwa!
we are launching muse in japan today! 🇯🇵
it is now available—download it from your favorite app store!
(i hope you're ready for the anime memes 😇)
Jack Krawczyk retweeted
I’m hiring an intern to work closely with me at the intersection of Health, RL and agents at MSL 🧑⚕️
Directions I am very excited about are taking shape, with plenty left to explore.
I’d love to hear from curious people who enjoy turning open-ended ideas into things that work!
Jack Krawczyk retweeted
As promised, Muse Spark is being open sourced today.
This was our first base model we trained under the new MSL org, this open weight version has gone through a lot of careful additional training for open-weight release. It’s an efficient and capable model, try it out!
Jack Krawczyk retweeted
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
the only thing better than Muse Wonder Full is updated token efficiency in Muse Code to resultmaxx
Muse Code 1.4.8 is out and is focused on token efficiency.
- Auto-review and background agents now use a cheaper model when available, and long auto-reviews stop sooner instead of burning tokens.
- We've fixed a bug that routed a small number of users who had enabled contributor mode to the full-priced models.
5
1,234
Jack Krawczyk retweeted
muse wonder is super fast and token efficient ⚡️ ⚡️
give it a go!
The two top models on the Intelligence Index take about twice as long per task: Claude Opus 5.5 and Sonnet 5.5 (@AnthropicAI) vs. Muse Wonder (@Meta).
Muse Wonder finishes an average task in 6.8 minutes with a 52.4 score. Opus 5.5 takes 13.2 and Sonnet 5.5 takes 15.3, scoring 57.6 and 56.0. Muse Spark 1.3 and GPT-6 Astra (@OpenAI) follow at 8.1.
When someone is waiting on the answer, minutes can matter more than a few index points. Merge Gateway lets you send latency-sensitive requests to the fast model and the hard ones to the strong one, behind one endpoint.
Scores via artificialanalysis.ai, week of Oct 12.






















